In this research work, four classifiers are adopted, analyzed, and discussed. The classifiers are Naïve Bayes (NB), Support Vector Machine (SVM), Stochastic Gradient Descent (SGD), and Logistic Regression (LR). The classifiers are operated on a dataset with more than eight-thousands of instances. The dataset contains the users' reviews and their opinions about the quality of service of restaurants. The reviews are collected from the Arabic Facebook posts. Several experiments are done to evaluate the performance of the adopted classifiers. Moreover, some features selection methods are also applied to improve the classification process. The feature selected methods are based on term-weights with N-grams, correlation, chi-square, and mutual information. Some criteria are considered to evaluate the performance of the classification process mainly: precision, recall, F-measure, and learning time. From the experimental results, the SVM classifier outperforms the other adopted ones. Also, the feature selection method based on the correlation between the individual features and the target class outperforms the other chosen methods. The same concluding remarks are expected to take place for other datasets containing comments or reviews from social media.